Aims: To identify relevant variables that may predict pain after routine extraction of erupted teeth, to construct a Pain Prediction Index (PPI) based on these variables, and to verify how these variables are related by using valid structural equation modeling (ie, path analysis). Methods: This study was designed as an observational prospective study for postoperative memory of pain after dental extraction of erupted teeth. Data from a total of 781 surgical procedures related to dental extractions were included. Pain was self-reported by the patients and was evaluated by a verbal category scale (VCS) on the seventh postoperative day. The database was searched for predictive variables that were significantly (P .05) associated with postoperative pain. Results: Pain was scored by patients as none in 65.4% of cases (511); light in 22.9% (179); moderate in 11% (86); and severe in 0.6% (5). Seven predictive variables were strongly related to postoperative pain: gender (female); age (younger than 33); number of extracted teeth (three); surgical technique (surgical flap, ostectomy, or teeth sectioning); number of local anesthetic cartridges (more than three); time in surgery (greater than 25 minutes); and any surgical complications. These variables were used to compose the PPI. The retained PPI classified 66.3% of the patients correctly for both pain and no pain; however, when more than three points were scored on the PPI, it correctly predicted pain in 55.6% of cases. Conclusion: The present study identified seven predictive variables that were strongly related to postextraction pain and documented that the PPI could correctly predict pain in the majority of patients.